The prevention and management of invasive fungal infections are of enormous importance in the care of patients undergoing cancer chemotherapy or bone marrow transplantation. Fungal infections such as invasive aspergillosis and candidiasis continue to be important causes of mortality in these patients. Key management strategies include accurate and rapid diagnosis, optimal use of antifungal drugs for prophylaxis and empiric therapy, and, in select patients, additional interventions such as immunomodulation therapies and surgery. This article will evaluate the newer antifungal agents, along with some agents that are still being investigated, for use in specific patient populations and for management of specific mycoses. Although not uniformly successful in every patient, these new agents nevertheless will have a very positive impact on the management of difficult invasive mycoses. Overview Some of the major developments in the treatment of invasive mycoses can be highlighted by a brief review of three landmark, evidence-based studies:

• Combination therapy for cryptococcal meningitis: van der Horst et al[1] compared amphotericin B(Drug information on amphotericin b) (0.7 mg/kg/d) with a combination of amphotericin B and flucytosine(Drug information on flucytosine) (100 mg/ kg/d) for the treatment of cryptococcal meningitis in patients with AIDS. After 2 weeks of treatment, cerebrospinal fluid (CSF) cultures were negative in 60% of those who received combination therapy, compared with 51% of those who received amphotericin B alone (P = .06).
• Voriconazole vs amphotericin B for invasive aspergillosis: Herbrecht et al[2] compared voriconazole(Drug information on voriconazole) at two doses (6 mg/kg on day 1, followed by 4 mg/kg twice daily) with amphotericin B deoxycholate (1 to 1.5 mg/kg/d) in patients with invasive aspergillosis. Successful outcomes were achieved in 52.8% of the voriconazole group, compared with 31.6% of the amphotericin B group.
• Caspofungin vs amphotericin B for invasive candidiasis: Mora-Duarte et al[3] compared the echinocandin caspofungin(Drug information on caspofungin) with amphotericin B in patients with invasive candidiasis. In a secondary analysis, caspofungin's success rate was approximately 15% higher than the success rate for amphotericin B. In addition, caspofungin was associated with a lower incidence of adverse events. It is important to note that the treatment success rates achieved in these robust, large, comparative studies clearly do not approach 100% (Table 1). Although the measurement of results in some clinical trials might be debatable because treatment success was based on specific study criteria, overall, the failure rates are still considerable, regardless of the clinical criteria used.

Diagnosis Considering the significant morbidity and mortality associated with invasive fungal infections in immunocompromised patients, it is particularly important to diagnose a fungal infection as early as possible. Unfortunately, the diagnosis is often challenging. In invasive aspergillosis, for example, the signs and symptoms are nonspecific, and blood cultures are rarely positive. Although it is often difficult to obtain samples for histopathology and culture, it is important to make every attempt to identify the organism because the findings will influence treatment. The combination of histopathology and culture is necessary because the results of one method might not be definitive. For example, regarding the technique of histopathology, the presence of branching septae and hyphae could indicate Fusarium, Aspergillus, or Scedosporium species, and only a culture would identify the correct pathogen. In addition, with histopathology, some patients may have conidia in tissue that produce adventitial forms of Fusarium, Acremonium, and Paecilomyces. The histopathology report may indicate that one of those organisms is the likely pathogen. In contrast, the finding of conidia and hyphae in tissue is not seen with Zygomycetes, Aspergillus fumigatus, or Aspergillus flavus. Thus, the histopathology results can help complement culture results. Although there is no absolute proof that early diagnosis results in less burden of organisms, there is some evidence that early diagnosis affects outcome. One study of patients with pulmonary aspergillosis indicated that more frequent use of bronchoscopy and high-resolution CT led to a more rapid diagnosis and, therefore, earlier treatment.[4] The mortality rate was 41% when the diagnosis was made within 10 days or less after the onset of signs and symptoms, compared with 90% when the diagnosis was made after 10 days. CT scanning can be very useful in the diagnosis of invasive pulmonary aspergillosis. The most characteristic CT findings are the halo sign and aircrescent formation. The halo sign is visualized as an area of low attenuation around a nodule or pleural-based lesion. Although these signs can occur in other diseases, they are highly suggestive of invasive pulmonary aspergillosis in febrile neutropenic patients. The halo sign has a short duration; therefore, the use of early CT can be quite valuable. A more nonspecific but common finding is the presence of nodular lesions on a CT scan; this has less predictive value. The diagnosis of invasive aspergillosis can be confirmed by serologic tests to detect circulating antigens, such as cell wall galactomannan and measurement of β-1,3-D glucan concentrations. For example, Kami et al[5] evaluated the sensitivity and specificity of real-time polymerase chain reaction (PCR), enzyme-linked immunosorbent assay (ELISA) for galactomannan, and the β-D glucan test for the diagnosis of invasive aspergillosis (Table 2). The study included 33 patients with invasive aspergillosis and 89 controls. The sensitivities and specificities ranged from 58% to 79%, and 84% to 97%, respectively.[5] Real-time PCR was more sensitive than the other two tests; it was also highly specific for Aspergillus infection. This same study also evaluated the relationship between test results and CT findings. Positive findings on PCR preceded those of CT by -0.3 ± 6.6 days; the corresponding figures were 2.8 ± 4.1 days for ELISA and 6.5 ± 4.9 days for the β-D glucan test. In high-risk patients, it is particularly important to use every tool available- whether it is a serologic test or CT scan-to identify these invasive fungal infections as soon as possible. Often, CT scans show a small nodule, which the radiologist interprets as po- tentially being fungal-related. In this situation, serologic findings could help support the diagnosis of aspergillosis. Although these tests are not perfect, using a combination of them in highrisk patients may result in an earlier diagnosis. The diagnosis of invasive candidiasis could be improved by better serologic tests, since at present, blood cultures are positive in only 50% to 60% of patients with invasive candidiasis. Immunomodulation Although immunomodulation has been extensively studied at the basic science level, clinically, its use has yet to be optimized. Some studies indicate that immunomodulation therapies do not improve outcome in immunocompromised patients. For example, a large, multicenter, European Organization for Research on Treatment of Cancer (EORTC) study demonstrated that the use of growth factors-primarily granulocyte colony-stimulating factor (G-CSF)- had no impact on outcome in hematology patients who had invasive aspergillosis.[ 6] This study included 130 cases, 20 hospitals, and 8 countries. Safdar et al,[7] at the M. D. Anderson Cancer Center, evaluated treatment with G-CSF-primed white blood cells (WBCs), given every other day, in 29 patients with candidemia; 76% of the patients were neutropenic. The control group consisted of 441 cancer patients with candidemia, 45% of whom were neutropenic. The associated mortality was 48% in patients who received WBC transfusions and 45% in the control group. The authors suggested that patients who received the WBC transfusions may have had worse prognostic factors than the control patients, and therefore, their comparable mortality may have actually represented a favorable response, but this interpretation will require further validation. Another approach to immunomodulation involves immune reconstitution. In a study by Pappas and associates,[ 8] two doses of adjuvant recombinant interferon-gamma were given to patients with cryptococcal meningitis. A trend toward faster decline in CSF yeast counts was seen in those patients who received the interferon, compared with those who received placebo. It is likely that the results were not statistically significant because there were not enough patients in the study. However, interferon may be difficult to use in organ transplant recipients because of the rejection risk to the transplanted organ. A word of caution: the use of growth factors to treat pulmonary aspergillosis has the potential complication of acute respiratory distress syndrome. This may be related to the massive amount of degranulation of WBCs in the area of the infection. There is some evidence that boosting the WBC count too fast is detrimental. For example, in a small study, the immediate mortality rate was 50% in patients who received G-CSF for aspergillosis who had a rapid increase in WBC count (0 to 4,500/μL in less than 5 days), compared with 17% in those who had a more gradual increase in WBC count.[9] It is clear that we need further studies to optimize our use of immunomodulation as adjunctive therapy for successful management of fungal infections. Antifungal Therapy: Drug Regimens or Prescriptions A number of key questions remain about antifungal drug regimens. The following is a representation of just some of these questions:

• What is the best dosage for liposomal amphotericin B? Ruiz et al[10] retrospectively studied 13 patients with proven, probable, or possible in- vasive aspergillosis. Four of the patients had received prophylaxis with itraconazole(Drug information on itraconazole), and eight had neutropenia. Treatment with 10 mg/kg/d of liposomal amphotericin B was successful in 9 of 10 patients with proven or probable disease, and none of the patients had nephrotoxicity. This finding does not prove that 10 mg/kg/d is the best dosage, but it suggests that the possibility that the higher doses might be beneficial should be considered.

• Are the dosages of echinocandins too low? This is a concern, particularly in neutropenic patients who have high relapse rates of candidemia. Schranz et al[11] evaluated the efficacy of three different doses of anidulafungin, an investigational glucan synthesis inhibitor, in the management of candidemia. The success rate was 72% for the lowest dose (100 mg loading dose/50 mg daily dose), 87% for the intermediate dose (150 mg/75 mg), and 87% for the highest dose (200 mg/100 mg). This result may reflect dose-dependent efficacy. Echinocandins have dose-dependent killing in vitro. In neutropenic patients, the organisms have to be killed primarily by the drug, because the host defense is limited. Since the echinocandins are relatively nontoxic, it is reasonable to ask whether we are underdosing when we give 50 mg of caspofungin.
• Is there a role for the aerosolized route of administration? There is a moderate amount of experience with aerosolized amphotericin B lipid complex (ABLC). Pharmacokinetic data suggest that the lipid formulations in suspension might be more effective than the deoxycholate preparation of amphotericin B. Certainly, the lipid products are much easier to aerosolize. In an open trial in which 51 lung transplant recipients received aerosolized ABLC, the incidence of toxicity was less than 5%.[12] There were no lung infections, two anastomosis infections, and four extrapulmonary infections, which you might expect to occur simply because this drug is not absorbed into the systemic circulation. In a randomized double-blind study, 100 consecutive lung transplant recipients were given aerosolized ABLC or aerosolized amphotericin B deoxycholate.[13] The incidence of adverse events was lower in the ABLC group than in the amphotericin B deoxycholate group (13.7% vs 28.6%; P = .03). Failures of prophylaxis occurred in 11.8% of the ABLC group and in 14.3% of the amphotericin B deoxycholate group. More studies are needed to determine whether fungal infections in the lung can be prevented with minimal amounts of inhaled antifungal drug, particularly when the drug needs to be given for long periods, as is the case with bone marrow transplant (BMT) recipients.
• To what extent does the site of infection influence the efficacy of specific therapies? CNS fungal infections remain quite difficult to treat. Troke et al[14] reported a 34% complete or partial response rate in 86 patients with CNS aspergillosis who were given voriconazole. Among BMT recipients, the success rate was 15%, while among the other patients, the success rate was 42% to 50%. Scedosporium prolificans and Scedosporium apiospermum can also cause CNS infection; the latter is relatively susceptible to voriconazole. In one study, voriconazole was effective in 7 out of 11 patients (64%) with S apiospermum infection but failed in 2 of 2 patients with S prolificans infection. Pitisuttithum et al[15] reported a 40% success rate with posaconazole in patients with CNS fungal infections, such as those caused by Aspergillus, Scedosporium, Coccidioides immitis, Histoplasma capsulatum, black molds, and Zygomycetes. Posaconazole's success rate was 59% in patients with cryptococcal meningitis.

Surgery The role of surgery for the management of invasive fungal infections should not be overlooked. Surgery does have a role in select patients with Zygomycetes and black mold infections and in some patients with invasive infections with Aspergillus and other hyalohyphomycetes. Caillot et al[16] reported that combined medical- surgical therapy achieved a cure rate of 84% in 25 neutropenic patients who had surgically proven pulmonary aspergillosis. Another study involved 87 patients with hematologic malignancies in whom invasive pulmonary aspergillosis was suspected.[17] Of the 39 patients who underwent resection on the basis of CT findings, 35 actually had invasive pulmonary aspergillosis. The 2-year survival rate was 36% in the resected group and 20% in the unresected group. Thus, there is evidence that in some circumstances, surgery can have a positive outcome, although it is important to note that these are very select patient populations regarding the ability to tolerate surgery. Antifungal Therapy: A Focus on the Newer Drugs Invasive fungal infections continue to be a significant cause of morbidity and mortality among patients undergoing cancer chemotherapy or bone marrow or stem cell transplantation. The toxicity associated with amphotericin B deoxycholate and the incidence of treatment failures have led to the development of new antifungal drugs, such as the lipid formulations of amphotericin B, the triazoles, and the echinocandins. Liposomal Amphotericin B
The lipid formulations of amphotericin B are approved for use in the treatment of invasive fungal infections when amphotericin B deoxycholate fails or when it is associated with unacceptable toxicity. Liposomal amphotericin B has been reported to have an overall response rate of about 60% in this setting[ 46,47] and is associated with less nephrotoxicity than amphotericin B deoxycholate.[48,49] Triazoles
Voriconazole is a broad-spectrum triazole with efficacy against invasive aspergillosis,[ 2,50] fluconazole(Drug information on fluconazole)-resistant candidiasis,[51] and a variety of other mycoses.[52] It is also effective in preventing breakthrough fungal infections in patients with fever and neutropenia. Voriconazole has been approved by the FDA for primary treatment of acute invasive aspergillosis and for salvage therapy for serious infections caused by Scedosporium apiospermum and Fusarium species. Data from 85 patients who received voriconazole have been reviewed.[53] Approximately 50% of the patients were bone marrow transplant recipients, approximately 27% had aspergillosis, and 24% had neutropenia; 13% received combination therapy. Within 72 hours of initiating voriconazole therapy, the most common drug interactions were with cyclosporin A (52%) and tacrolimus(Drug information on tacrolimus) (29%); for both, the levels fell outside the therapeutic range. Liver function test results were elevated in 10% of patients. Therefore, when using voriconazole, it is important to be aware of these drug interactions and to carefully consider the cost-benefit ratio. Echinocandins
The echinocandins include caspofungin, micafungin(Drug information on micafungin), and anidulafungin. These drugs are glucan synthesis inhibitors. Caspofungin is active against Aspergillus and Candida species, but it does not have significant activity against Cryptococcus neoformans. It has good in vitro activity against Candida species, including those resistant to fluconazole and itraconazole.[54,55] Caspofungin has been demonstrated to be effective in the treatment of oropharyngeal and esophageal candidiasis,[56-58] fluconazole-resistant esophageal candidiasis,[59] and invasive candidiasis.[3] Caspofungin is approved by the FDA for salvage therapy in patients with invasive aspergillosis who have been refractory to or intolerant of amphotericin B, amphotericin B lipid complex, and/or itraconazole. It is also approved for the treatment of oropharyngeal and esophageal candidiasis and invasive candidiasis. The standard dosage is 50 mg/d IV after a 70-mg loading dose; higher dosages (70 mg/d) have been safely used. The duration of treatment depends on the severity of the patient's underlying disease, recovery from immunosuppression, and clinical response. Caspofungin has an excellent safety profile and does not appear to be antagonistic when combined with other antifungal therapies.[60] Micafungin is an investigational agent that has a broad spectrum of activity against Candida species[61] and Aspergillus species. Anidulafungin is also an investigational agent that is active against Candida species.[62,63]